A green backup job tells you a file was written. Only a restore tells you the business can run from it, and most firms find out the difference on their worst day.
Backups fail quietly. A job skips an open database, a retention setting shortens without anyone noticing, or the copy sits on the same network as the thing it protects. None of this shows on a dashboard. All of it shows up during a restore.
A small test each quarter
Pick one system each quarter and bring it back somewhere safe. Time it, and write down what was missing.
- One system, chosen by the business, not by IT.
- A clean network, away from production.
- A stopwatch and a note taker.
- One page of findings at the end.
Nobody has ever been thanked for a backup. Only for a restore.
What you will find
Expect missing licences, undocumented passwords and a restore time longer than anyone guessed. That is the point of the exercise. Each one is cheap to fix on a quiet Tuesday and very expensive during an incident.
In a sample city report, records came back in 31 hours because an offline copy existed and people knew the order to bring systems up. The order mattered as much as the copy. A file server is no use if the directory it depends on is still down.
Three questions for your next review
Put these to whoever owns backup. If the answers come with dates, you are in good shape.
- When did we last restore something the business relies on?
- Is there a copy an attacker with admin rights could not delete?
- Who decides the restore order, and where is it written down?
Write the time from each test into the runbook. Next quarter, you will have a real number to give the board in place of a hopeful one.




